NEWS.md
combine_supp() requires that the QNAM columns are not in the source dataset (#64)build_from_derived() no longer requires all datasets with predecessors, but can run on a single dataset or multiple. It will also now accept unnamed datasets.combine_supp() now allows multiple QNAM values to go to the same IDVAR (#63)combine_supp() (#45)check_ct_data() so that na_acceptable can now accept a vector of variables. Also add new argument omit_vars to control if any variables should be skipped when doing controlled terminology checks (#57)check_unique_keys() to check the uniqueness of records in the dataset by variable keys (#52)build_from_derived() predecessor_only parametercombine_supp() when the data and the supp have white space. Now it will be trimmed before attempting to mergebuild_from_derived() that prevented multiple from being combinedfloating_pt_correction from combine_supp() as it is best to never change the type of the main datasetdrop_unspec_vars to explain which variables are droppedorder_cols, so it will still work when not all columns are presentcombine_supp() when the data and the supp have different classes for the IDVARVALcombine_supp() to report when not all the rows in the supp have mergedfloating_pt_correction argument to combine_supp() used for when there are floating point errors with IDVARVAL